Climate Change and Geohazards in South West England

Climate change and geohazards in South-West England
CLIMATE
CHANGE AND GEOHAZARDS IN SOUTH-WEST
PROJECTIONS, PROGRESS AND CHALLENGES
ENGLAND:
E.J. PALIN
Palin, E.J. 2012. Climate change and geohazards in South-West England: projections, progress and challenges.
Geoscience in South-West England, 13, 31-46.
The Earth’s climate has varied naturally throughout geological time, but there is very strong evidence that anthropogenic activity
since the Industrial Revolution has superimposed a further climate change component onto the background of this natural climate
variability and change. In this paper, the causes of climate change - both natural and anthropogenic - are summarised. The models
which are used to simulate the climate, and to make projections of future climate change, are described, and the sources of
uncertainty in these projections are also discussed. Recent projections of changes in temperature, precipitation and sea level are
outlined, at global, UK and South-West England scales. Finally, some potential links between climate phenomena and
geological/geomorphological hazards in South-West England are investigated, highlighting a selection of the current challenges in
climate impacts research and the science of climate risk.
Met Office Hadley Centre, FitzRoy Road, Exeter EX1 3PB, U.K.
(E-mail: [email protected]).
Keywords: Climate change, climate impacts, UKCP09, South-West England, geohazards, risk.
INTRODUCTION
Climate change is a cross-cutting societal issue, with
potential implications for everyone. Climate scientists develop
the underpinning science and computer modelling capability,
which together permit the simulation of the climate system,
thereby furthering our understanding of the climate and
allowing projections of future climate change to be made.
Politicians and policy-makers, informed by these projections,
are typically then involved with coordinating a policy response
to climate change at a variety of administrative levels
(international, national, local, etc.). Finally, either directly in
response to projected changes, or indirectly in response to
policies, stakeholders - such as industry and the general public
- can then consider how they might manage the potential
impacts of climate change on their day-to-day activities and
lives.
One of the most well-known organisations with respect to
climate change is the Intergovernmental Panel on Climate
Change (IPCC), whose Fourth Assessment Report (‘AR4’) was
published in 2007. This was the most recent in a series of
assessment reports synthesising the state of scientific
understanding of climate change. The first volume (Working
Group I, ‘AR4 WG1’ (IPCC, 2007b)) described the physical
science basis for climate change. Whilst it is clearly not
possible to discuss all the findings of AR4 WG1 in this paper,
amongst the conclusions was that “warming of the climate
system is unequivocal, as is now evident from observations of
increases in global average air and ocean temperatures,
widespread melting of snow and ice, and rising global average
sea level ”.
Climate change per se is not unusual; from a geological
perspective, it is known that the climate has varied in the past.
In AR4 WG1, evidence from various palaeoclimatic proxies
(including isotope ratios in sediments and seawater) was
examined, and from this it was concluded that “pre-Quaternary
climates prior to 2.6 Ma were mostly warmer than today”
(Jansen et al., 2007). More recent palaeoclimatic proxy data,
from ice cores, resolve the several glacial-interglacial cycles
which have occurred over the past ~700 ka (Solomon et al.,
2007). However, considering the (geologically) very recent
past, the climate change that has occurred is considered
unusual. AR4 WG1 notes that “palaeoclimatic information
supports the interpretation that the warmth of the last half
century is unusual in at least the previous 1300 years […]
Average Northern Hemisphere temperatures during the second
half of the 20th century were very likely (>90% probability)1
higher than during any other 50-year period in the last 500
years and likely (>66% probability)1 the highest in at least the
past 1300 years” (Solomon et al., 2007). Although there has
been natural variation of the climate in the geological past,
there is very strong evidence that human activity has been a
driver of climate change since the Industrial Revolution.
This paper begins with an overview of the major natural and
anthropogenic (human-induced) causes of climate change.
Next, the computer models used to produce climate change
projections are described and the major sources of uncertainty
in those projections are examined. Following this, examples
from the most recent set of climate change projections for the
UK, and South-West England in particular, are discussed.
Approaches to acting upon projected changes to the climate are
also considered. Finally, some of the current challenges
associated with climate modelling and climate impacts research
are explored, using case studies of past events in South-West
England, and linking climate change with other
geological/geomorphological hazards.
31
E.J. Palin
THE
CLIMATE SYSTEM AND NATURAL CAUSES OF
CLIMATE CHANGE
The climate system consists of the atmosphere, ocean,
biosphere, cryosphere (ice and snow) and geosphere (rock and
soil). The internal interactions and feedbacks between these
components, plus factors external to the system, result in
variations in climate over many different timescales and length
scales, even without any anthropogenic influence. Below we
consider the internal variability of the climate, together with the
three major natural external causes of climate change: variation
in (a) the Earth’s orbit around the sun, (b) solar output, and (c)
atmospheric aerosol from volcanoes.
Internal climate variability
The internal variability of the climate results from the chaotic
nature of the climate system, characterised by the ‘butterfly
effect’ (Lorenz, 1963) embodied by the idea of “the flap of a
butterfly’s wings in Brazil setting off a tornado in Texas”2, which
describes the sensitivity of the climate system’s behaviour to the
initial conditions - that is, a small change at one location in a
non-linear system can result in large differences to a later state
of the system. Implicit in this analogy is the spanning of many
different scales in space and time, and indeed such variability
is evident in weather and climate phenomena: from individual,
relatively short-lived and small-scale storms to large-scale, selforganising modes of climate variability. Examples of the latter
include (i) the El Niño-Southern Oscillation, a major climatic
oscillation with a period of 2-7 years resulting from oceanatmosphere coupling in the tropical Pacific, which affects the
climate of many parts of the globe, and (ii) the North Atlantic
Oscillation, a variation in pressure patterns in the North Atlantic
which chiefly affects the climate of Europe (Solomon et al.,
2007).
In the presence of a climate change ‘signal’, the ‘noise’ of
internal variability could either reinforce or oppose the signal in
a given place at a given time. Internal variability is one reason
why, for example, the occurrence of one cold winter does not
signify the cessation of global warming. It is also important to
note that climate change could itself also alter the characteristics
of internal variability.
External climate variability
Orbital variations: Over the period of many thousands of
years, the Earth’s orbit around the sun changes in a quasiperiodic manner. There are three principal cycles, called
‘Milankovitch cycles’ after the Serbian mathematician who first
described them in mathematical detail (Milanković, 1941)3,
which describe the variation in the shape (eccentricity) of the
orbit, and in the axial tilt (obliquity) and axial precession. Here,
we restrict the discussion to these principal variations, but note
that others do exist. The periods of the principal variations are
approximately 100 ka, 41 ka, and 19-23 ka respectively
(Solomon et al., 2007). These cycles affect the total amount of
solar radiation incident upon the Earth’s upper atmosphere
(‘total solar irradiance’), and also the extent of seasonal and
latitudinal variations in solar irradiance (‘insolation’). Hays
et al. (1976) were the first authors to demonstrate a correlation
between climatic oscillations (specifically between glacial and
interglacial periods) and Milankovitch cycles; there is
considerable evidence in the form of palaeoclimatic analysis of
Antarctic ice cores, which also show strong cyclic
(glacial/interglacial) variations over the last few hundreds of
thousands of years (e.g. Petit et al., 1999; EPICA community
members, 2004).
Solar output: In recent decades it has been possible to use
satellites to measure the radiation output by the Sun, which
revealed slight variations in solar output (Le Treut et al., 2007).
This, in addition to the orbital variations described above,
is another factor influencing total solar irradiance. The data
32
thus observed show a well-established cycle with a period
of approximately 11 years; other cycles with longer periods
have also been postulated to exist. The main known cause of
present-day irradiance variability is the presence of sunspots
and faculae on the Sun’s disc: respectively, dark regions
with locally depleted radiation, and bright regions with locally
enhanced radiation (Solomon et al., 2007). Larger numbers of
sunspots correspond with higher irradiance, and vice versa.
Although solar variability in the satellite era is wellcharacterised, reconstruction of solar irradiance before this
period requires the use of proxies including historical records
of sunspot numbers and cosmogenic isotope measurements;
such reconstructions are prone to greater uncertainty. For
example, the historical record of sunspots extends back over
2000 years (Wittmann and Xu, 1987), but only over the last four
centuries (Vaquero, 2007), since the invention of the telescope,
have there been sufficient data to serve as a reliable proxy. A
major feature of the record since this time is the ‘Maunder
minimum’, a period of low solar activity (and hence
comparatively few sunspots), between 1672 and 1699 (Le Treut
et al., 2007). This time corresponds to part of the climatic
period known as the ‘Little Ice Age’ (LIA), when some regions
of the world are thought to have sustained temperatures
notably cooler than the rest of the last millennium, but the issue
of any causal relationship between solar activity and
temperatures during this period is still a matter of debate, as is
the extent to which any recently-observed changes in climate
could be related to solar activity.
Some examples relevant to 20th Century climate include the
study by Scafetta and West (2006) which stated that solar
variability has a significant effect on climate forcing, and that
the Sun might have contributed up to 45-50% of the 1900-2000
global warming, and 25-35% of the 1980-2000 global warming.
Another study using satellite observations concluded that recent
solar variations are too small to have contributed appreciably to
the warming which has been observed over the last three
decades (Foukal et al., 2006). In addition, a review by
Lockwood and Fröhlich (2007) found that “over the past 20
years, all the trends in the Sun that could have had an influence
on the Earth’s climate have been in the opposite direction to that
required to explain the observed rise in global mean
temperatures”. It should be noted that solar radiation effects
are further complicated by variations across the spectrum, i.e.
not just the gross radiation, in any effects on climate. For
example, a link has been postulated between ultraviolet
irradiance variability and the winter climate of the Northern
Hemisphere (Ineson et al., 2011). Solar influence on the
climate continues to be an area of active research.
Atmospheric aerosol from volcanoes: The term ‘aerosol’
refers to any particulate matter suspended in the atmosphere
and includes natural and anthropogenic contributions (the latter
are discussed below). Through a variety of mechanisms
(Forster et al., 2007)4, some better understood than others,
aerosols reduce the amount of solar radiation arriving at the
Earth’s surface, thereby exerting a net surface cooling effect. In
the context of external climate variability, large volcanic
eruptions are one contributor to the natural atmospheric aerosol
burden. The eruption of Mount Pinatubo in 1991, for example,
allowed scientists to assess the climatic impact of a major
eruption. Around a year after the eruption, reductions in global
mean air temperature of around 0.5°C were observed at the
surface (Parker et al., 1996).
There are other natural atmospheric aerosols, including dust,
for example from deserts5, dimethylsulphide (DMS) produced
chiefly by marine algae, and salt, especially in the marine
atmosphere. However, although these are natural in origin,
here they are not considered together with volcanic aerosol
because while volcanoes are driven by processes outside the
climate system, and as such contribute to external climate
variability, this is not the case for the remaining natural
aerosols.
Climate change and geohazards in South-West England
ANTHROPOGENIC
CAUSES OF CLIMATE CHANGE
Despite the fact that the Earth’s climate has always varied
naturally, there is strong evidence that anthropogenic influence
on the climate system is also affecting the climate. This is
demonstrated in two key points from AR4 WG1 (IPCC, 2007c):
1)
“Most of the observed increase in global average
temperatures since the mid-20th century is very likely
(>90% probability)1 due to the observed increase in
anthropogenic greenhouse gas concentrations”.
2)
“It is likely (>66% probability)1 that increases in
greenhouse gas concentrations alone would have
caused more warming than observed because volcanic
and anthropogenic aerosols have offset some warming
that would otherwise have taken place”.
These two statements also make reference to the two major
anthropogenic causes of climate change, which are (a)
anthropogenic greenhouse gases and (b) anthropogenic
atmospheric aerosol.
Anthropogenic greenhouse gases
Greenhouse gases (GHGs) are those gases which, due to
their molecular structure, absorb and emit radiation at specific
wavelengths within the spectrum of thermal infrared radiation
emitted by the atmosphere, the Earth’s surface, and clouds
(IPCC, 2007b). The greenhouse effect, by which heat is
‘trapped’ in the lower atmosphere as a result of the presence of
these gases, is a natural effect, without which the Earth’s surface
would be around 33°C cooler (Le Treut et al., 2007) and hence
far less conducive to sustaining life.
However, the
anthropogenic increase in GHGs since the Industrial
Revolution, from processes such as land-use change and
burning fossil fuels, is resulting in an enhancement of the
greenhouse effect - in addition to the effect which would be
experienced in the absence of human influence – and it is this
anthropogenic enhancement that is contributing to unusual
changes in the climate.
The principal anthropogenic GHGs are carbon dioxide CO2,
methane CH4 and nitrous oxide N2O. Water vapour is a major
GHG, but it is not listed here, because anthropogenic activity
does not have a large effect on water vapour concentrations in
the atmosphere. The main role of water vapour in connection
with anthropogenic GHGs is to respond to and amplify their
effects through water vapour feedback.
Increased
concentrations of the anthropogenic GHGs lead to a warmer
atmosphere, which in turn can hold more water vapour per unit
volume (according to the Clausius-Clapeyron relation). Since
water vapour behaves as a GHG, this results in further
warming, and hence a positive feedback occurs (Le Treut et al.,
2007).
Anthropogenic atmospheric aerosol
There are also anthropogenic sources of aerosol. These
include black carbon aerosol, from sources such as biomass
burning; chemical aerosol such as sulphates and nitrates, from
industrial sources; and the desert dust derived from humaninduced desertification5. The climatic effects of anthropogenic
aerosol are similar to those of natural aerosol, though there is
variation in the geographic distribution of the proportion of
aerosol derived from natural and anthropogenic sources.
The net radiative forcing6 due to atmospheric aerosol (both
natural and anthropogenic) is most likely negative. Positive
radiative forcing leads to a global mean surface warming, while
negative radiative forcing leads to a global mean surface
cooling (Forster et al., 2007). The likely negative radiative
forcing of atmospheric aerosol has led to its consideration as a
geoengineering solution, i.e. for use in deliberate manipulation
of the climate in order to try to offset the effects of warming.
In this case the aerosol would be deliberately injected into the
atmosphere at appropriate locations and heights. However, a
major study into geoengineering the climate (Royal Society,
2009) recommended caution with respect to the potential
implementation of this and many other geoengineering
proposals, since they are not well-understood in terms of their
impacts, efficacy and cost. The study also noted explicitly that
geoengineering should not divert policy away from other
strategies for managing future climate change. The two main
classes of such strategies will be briefly discussed later in this
paper.
CLIMATE
MODELS
In order to produce projections of future climate, scientists
use climate models. These are mathematical models which
encapsulate within a set of equations the essential physics of
the climate system.
These equations are solved, at an
appropriate resolution, in space and time. As with most types
of computer simulation, early climate models were relatively
primitive and had coarse resolution. Nowadays, improvements
to our understanding of physical processes affecting the
climate, coupled with huge increases in computational capacity,
have resulted in present-day climate models which can be both
scientifically complex and also have comparatively fine spatial
and/or temporal resolution (Le Treut et al., 2007). That said,
while increasing complexity and/or resolution is generally
considered advantageous for many purposes, simpler models
still have a role to play. Indeed, the ‘hierarchy of models’ is
considered to be a useful mechanism in linking theoretical
concepts through to the increasing ‘realism’ of the more
complex models (Held, 2005). As time progresses, climate
models become ever more sophisticated, and their developers
continually assess their performance. Typically this is achieved
by examining how well the models are able to reproduce key
elements of past climate and past climate change, when driven
by observed historical forcings.
Climate models are not perfect predictors and our
understanding of the climate system is not perfect. There are
processes which are known but not well-represented in
models, and it is also likely that there are aspects of the climate
system of which we are currently not aware, and which are
hence not represented at all in models. As well as these factors,
there are other assumptions made regarding climate model
input, as discussed below. Therefore, although very significant
scientific progress in the area of climate modelling has been
made in recent years, there is still much to be done, and many
uncertainties to tease out (Hawkins and Sutton, 2009).
UNCERTAINTY
IN CLIMATE CHANGE PROJECTIONS
Climate change projections are subject to a range of inherent
uncertainties originating from a variety of sources. The main
types of uncertainty are (a) natural variability, (b) forcing
uncertainty, and (c) uncertainty in the response of the climate
system (often termed ‘modelling uncertainty’). These terms,
together with some of the approaches used to try to manage
and/or understand the uncertainty in projections, are discussed
below. The term ‘ensemble’ will be used frequently in the
discussion. It denotes a group of parallel model simulations
used to produce climate projections, with different types of
ensemble being relevant for different applications.
Natural variability
The uncertainty due to natural variability is handled in
different ways depending on the scientific question being
addressed. For long-term projections (e.g. out to the latter half
of the 21st Century), we are typically interested in the ‘signal’
due to climate change, and want to separate out the ‘noise’ of
natural variability. Indeed, we cannot reliably predict whether
33
E.J. Palin
a given year will be warmer or cooler than average at such far
time horizons. Hence, to examine the climate change ‘signal’,
we can calculate the climate parameters of interest over
relatively long timescales, thereby smoothing out the interannual ‘noise’. Often, 30-year time periods are used for both
the baseline7 and future climate. Using several different future
periods and potentially more than one baseline period is also
common. However, this does not account for the possibility
that longer-term variations are in operation. There are aspects
of the climate that are known to vary on decadal or
multidecadal timescales, which may not be adequately
addressed by using 30-year periods.
With some modifications, climate models can also be used to
make nearer-term projections, such as those for the next decade
or so. In that case, we are actually interested in the variability
- that is, whether a particular year might be warmer or cooler
than average. The climate change trend, although present, is
relatively small compared to the variability. We can therefore
run a climate model many times with slightly different initial
conditions (an initial condition ensemble, ICE), enabling us to
estimate the spatial and temporal importance of the component
of the uncertainty which arises from natural variability. The Met
Office’s decadal prediction system, DePreSys (Smith et al.,
2007), uses this approach. A version of the climate model is
initialised with observed data for the ocean and atmosphere, as
well as plausible changes in other drivers of the climate
discussed above, and a range of different start dates is used,
thereby forming the ensemble.
Analogous runs without
assimilation of the observed data are also performed.
Comparing these sets of runs has shown that the assimilation
improves the forecast skill of globally-averaged surface
temperature throughout the decade (Smith et al., 2007).
Improving the resolution and skill of forecasts at these
timescales is a major research activity.
Forcing uncertainty
In order to produce climate projections, an estimate of the
future amounts of GHGs in the climate system is required as
input to climate models. However, we obviously cannot know
for certain what our real future GHG emissions pathway will be.
This issue has typically been addressed by producing climate
change projections under a range of emissions scenarios (e.g.
Nakićenović et al., 2000). These are plausible future pathways
which GHG emissions might follow, depending on factors such
as population, energy mix, technological advances, scales of
international cooperation, etc. In addition, the relationship
between GHG emissions and GHG concentrations in the
atmosphere is climate model-dependent (for example, CO2
partitioning between different parts of the climate system
depends on the representation of the carbon cycle in the
climate model).
Uncertainty in the response of the climate system
(often termed ‘modelling uncertainty’)
Although all climate models are based on the same realworld physics, different climate models have different
formulations and approximations according to the research
institutions where they have been developed. Processes
operating on a scale which is less than the grid resolution of the
climate model, such as the formation and evolution of clouds,
have to be parameterised, and these parameterisations will vary
from model to model. Additionally, although models are
constantly developed and updated to reflect new scientific
discoveries and understanding, there will always be some
potentially influential processes which are missing. A good
example is the representation of the carbon cycle. This is
clearly an important component of the climate, but it has taken
some time for it to be represented in many climate models.
C4MIP (Coupled Climate Carbon Cycle Model Intercomparison
Project), an intercomparison of 11 models with a ‘first34
generation’ interactive and parameterised carbon cycle, was
recently undertaken (Friedlingstein et al., 2006), but the models
involved have typically been especially constructed for carbon
cycle studies, rather than being the ‘mainstream’ model
developed by a particular climate research institution. It is only
recently that the ‘mainstream’ models have begun to include a
coupled carbon cycle, for example the Met Office Hadley
Centre’s earth system model, HadGEM2-ES (Jones et al., 2011).
Finally, the absence of particular processes may lead to further
uncertainties in the regional responses of climate models, as the
importance of a particular process (such as the monsoon, for
example) may be greater or smaller in a particular region.
There are various methods for addressing modelling
uncertainty, and typically a particular method will only address
a subset of the factors contributing to this uncertainty. Here we
discuss two of the major approaches. The first is use of a multimodel ensemble (MME) - that is a set of climate projections
comprising results from different climate models and/or
different research institutes. The IPCC AR4 MME comprised 21
different ensemble members; the differing degrees of consensus
between these models for different variables (e.g. temperature
and precipitation) will be briefly discussed below. Use of an
MME is usually an improvement on using just one model to
make a projection. However, one potential drawback of an
MME is that it is typically an ‘ensemble of opportunity’ - that is
a collection of model runs which happen to be available, rather
than having been subject to any particular scientific design
criteria. As such, the sampling is neither systematic nor random
(Tebaldi and Knutti, 2007). A related fact is that models from
which an MME is formed are not designed to sample the known
full range of uncertainty, because they are (not surprisingly)
tuned to match observational data as closely as possible. As
such, there is always the possibility that a model gives “the right
result for the wrong reason” (Tebaldi and Knutti, 2007). It has
also recently been suggested (e.g. Pennell and Reichler, 2011)
that, despite there being 21 ensemble members in the AR4
MME, the effective number of independent models is actually
considerably fewer than this, because of the similarities in
formulation between different models (e.g. the sharing of a
particular module of code between models, the development of
more than one model at the same institution, etc.).
The second approach is to use a perturbed physics ensemble
(PPE), which is a set of climate projections comprising results
from one climate model, but run many times with slightly
different input parameters. Expert judgement is used to
determine (a) which parameters may vary, (b) between which
values each parameter may vary and (c) which parameters may
co-vary in physically sensible ways. The PPE does not sample
the structural uncertainty (the fact that different models are set
up differently), but it does sample the parameter uncertainty
(the fact that our scientific understanding may not be sufficient
to know a particular parameter’s ‘true’ value in the real world,
or how this may evolve in future).
Technically, it is also possible to combine the MME and PPE
approaches, and a state-of-the-art suite of climate projections in
which there has been some incorporation of information from
the IPCC MME into a PPE will be discussed later in this paper.
However, combining MME and PPE approaches from first
principles is a demanding task, both from an organisational and
computational perspective, and hence has not yet been
undertaken comprehensively.
CLIMATE
CHANGE PROJECTIONS FOR THE
UK
We have already noted that there have been periods in the
geological past where the climate is thought to have been
warmer or cooler than the present (Jansen et al., 2007), but the
speed of the recent observed changes, and the projected future
changes, is surprisingly rapid. The projected change in global
average surface temperature by 2090-2099 (relative to 19801999) is between 1.1 and 6.4°C (IPCC, 2007c), depending on
the climate model used, emissions scenario considered, and
Climate change and geohazards in South-West England
strength of carbon cycle assumed, which is a projected change
of considerable magnitude over the course of just one century.
When driven by non-mitigation scenarios (i.e. those emissions
scenarios where no attempt is made to mitigate the effects of
climate change by for example reducing atmospheric GHG
concentrations), all models in the IPCC AR4 MME project
increases in global average surface temperature over the 21st
Century.
Projected changes in precipitation are for an increase in
globally-averaged precipitation in future, consistent with the
fact that a warmer atmosphere can hold (and thus release) more
water vapour. However, the projected regional variation in the
response for precipitation is less certain, with some regions
projected to experience an increase in precipitation, and some
a decrease. Agreement across climate model projections from
different centres is also lower for precipitation than for
temperature.
These very broad conclusions pertain to global climate
change and are generally arrived at using global climate models
(GCMs), whose resolution is relatively coarse. To assess climate
change on a regional basis, it is instructive to use climate
information at a higher resolution. In 2009 the most recent set
of climate projections covering the UK - UK Climate Projections
2009 (Defra, 2011), hereafter ‘UKCP09’ - was released. UKCP09
describes projected changes to the UK’s climate during the 21st
Century, using leading science developed at the Met Office
Hadley Centre. UKCP09 is a probabilistic set of projections,
into which some of the uncertainties associated with the climate
system have been systematically incorporated.
The
probabilities have been estimated by using a sophisticated
approach combining a large ensemble of climate model
projections, expert judgement, and statistical methods. The
result is that UKCP09 explores a large range of possible climate
futures and much larger than that of its predecessor, UKCIP02,
which did not follow this probabilistic approach. Projections
are available under three different emissions scenarios - low,
medium and high, corresponding to SRES B1, A1B and A1FI
respectively (Nakićenović et al., 2000) - and for seven 30-year
timeslices, from the 2020s to the 2080s (each named after their
middle decade, i.e. ‘2020s’ indicating 2010-2039, etc.), relative
to a baseline climate of 1961-1990.
UKCP09 was developed using the Met Office Hadley
Centre’s regional climate model (RCM), HadRM3 (Jones et al.,
1997), which has a resolution of 25 km. GCMs have global
coverage, but RCMs are typically run over a limited domain of
the Earth’s surface.
They therefore require ‘boundary
conditions’, which are data from the GCM at the boundaries of
the simulation domain, to drive them. For UKCP09, HadRM3
was driven by data from a version of the HadCM3 GCM
(Gordon et al., 2000). Information from other GCMs has also
been incorporated into the methodology, thereby sampling
some of the structural uncertainty, but only HadCM3 was used
to provide the RCM driving data. It is therefore reasonable to
assume that any systematic biases in HadCM3 could be
represented to some extent in UKCP09. Similarly, the extent to
which particular climate processes/constituents are modelled in
the driving GCM and the other models contributing to the
UKCP09 methodology will directly affect the way in which
those processes/constituents are represented in UKCP09
(Murphy et al., 2009). An example is the representation of
Atlantic blocking events in models. Atlantic blocking events are
high-pressure systems which impede the progress of
depressions along the Atlantic ‘storm track’. In the UK,
blocking can cause summer heatwaves and winter cold spells,
yet its representation in IPCC AR4-generation models (of which
HadCM3 is one) is known to be subject to systematic errors.
The simulation of blocking in climate models has only recently
begun to be improved (Scaife et al., 2011). Despite these
caveats, the UKCP09 methodology represents a state-of-the-art
approach to providing climate change projections in a format
which may be used to inform probabilistic risk-based planning.
The following are some of the conclusions quoted in the
UKCP09 briefing report (Jenkins et al., 2009). These are for
changes by the 2080s under the medium emissions scenario:
1)
All areas of the UK are projected to become warmer,
with greater projected warming in summer than winter.
2)
There is little evidence for a change in annual
precipitation amounts around the UK. However:
a) The largest projected increases in winter
precipitation are seen along the western side of the
UK. Decreases are projected over parts of the
Scottish Highlands.
b) The largest projected decreases in summer
precipitation are seen in parts of the far south of
England. Changes close to zero (ranging from a
slight decrease to a slight increase) are projected
over parts of Northern Scotland.
3)
Sea levels are projected to rise around the UK, with
larger rises in the south of the UK than in the north.
This geographical variation is due to differences in
vertical land movement as a result of glacial isostatic
adjustment since the last Ice Age.
For the UK, this summary shows that there is greater
uncertainty in the projected precipitation response (which may
be an increase or decrease) than that for temperature. This is
analogous with the global responses from the MME and also
with those from the global HadCM3 PPE, which was used to
drive the RCM simulations underlying UKCP09 (the latter is not
surprising as there will be a relationship between the global
PPE and regional PPE simulations because one is used to drive
the other). Note also that the briefing report summary above is
for one time period and one emissions scenario; other time
period/scenario combinations will give different projected
changes.
Probabilistic projections for temperature, precipitation,
humidity, cloud, pressure and atmospheric radiation fluxes are
available from the initial release of UKCP09. Subsequently,
various technical notes updating the initial release, and
probabilistic projections for wind speed, have been added
(UKCP09, 2012).
Here, only projections for changes in
temperature, precipitation and sea level are discussed to
illustrate the utility of UKCP09. Probabilistic results from
UKCP09 are typically quoted at the 10, 50 and 90% ‘probability
levels’ (though results for other levels are available).
Probability levels should be interpreted as indicating the
relative likelihood of the projected change being at or less than
the given change. To illustrate this concept, a representative
cumulative distribution function (CDF) for a projected
temperature change is shown in Figure 1. The 10, 50 and 90%
levels can be considered using the phrases ‘very unlikely to be
less than’, ‘central estimate’ and ‘very unlikely to be greater
than’, respectively. Hence, a projected change of +1.2°C at the
10% level, +2.3°C at the 50% level and +3.2°C at the 90% level
(Figure 1) can be described as “very unlikely to be less than
+1.2°C, very unlikely to be greater than +3.2°C, with a central
estimate of +2.3°C”. This kind of information is useful in terms
of risk-based planning. For example, a decision-maker may
choose to formulate their plans based on the central estimate or
the unlikely outcomes depending on their appetite to risk.
Other visualisation products besides the CDF are available
from UKCP09. These aim to represent the uncertainty in the
projections in a variety of different ways, so that the information
contained in the projections can be maximally useful to a range
of users with widely-varying requirements. More details can be
found in the UK Climate Projections Science Report (Murphy
et al., 2009).
35
E.J. Palin
Figure 1. Schematic cumulative probability distribution function
(CDF) for a projected temperature change, showing the concept of
‘probability levels’.
CLIMATE CHANGE
ENGLAND
PROJECTIONS FOR
SOUTH-WEST
A previous geological perspective on climate change
presented to the Ussher Society included a variety of climate
projections for South-West England (Hart and Hart, 2000).
Since UKCP09 provides projections aggregated to the level of
administrative regions, of which South-West England is one, an
updated set of region-specific results for South-West England is
presented here. It should once more be borne in mind that,
since the same methodology underlies the region-specific and
UK-wide projections, the caveats and uncertainties relevant for
the UK-wide projections will also be relevant for the South-West
England projections.
Projected changes to temperature and
precipitation in South-West England
Key findings for temperature and precipitation in South-West
England for the 2050s (under all three UKCP09 emissions
scenarios), with respect to a 1961-1990 baseline, are given in
Table 1. For the 2050s, under all three scenarios, there is a
projected increase in mean temperatures in both winter and
summer, and mean daily minimum and maximum summer
temperatures are also projected to increase.
In winter,
precipitation is projected to increase, whilst summer
precipitation projections are more uncertain; the mean change
is for a decrease, but the fact that the 90% probability level
shows a positive change means that an increase in summer
precipitation in this region cannot be ruled out. Once again it
should be stressed that these projected changes are for a
particular time period and projected changes for other time
periods will be different.
Projected changes to sea level around South-West
England
The three major processes affecting sea level are (a) thermal
expansion of sea water (with warmer water occupying a larger
volume than cooler water), (b) the exchange of water between
the oceans and other stores (e.g. glaciers and ice caps, ice
sheets, etc.), and (c) vertical land movement (e.g. glacial
isostatic adjustment, tectonic processes, etc.) (Bindoff et al.,
36
Table 1. Projected changes to climate parameters for South-West
England, under low, medium and high emissions scenarios, at a
range of probability levels by the 2050s, with respect to 1961-1990
values. Twm = mean winter temperature; Tsm = mean summer
temperature; Tsx = mean daily maximum summer temperature;
Tsn = mean daily minimum summer temperature, Pwm = winter
mean precipitation, Psm = summer mean precipitation. Source:
UKCP09.
2007). The first two of these affect the global ocean volume,
whilst the third changes the size and shape (and thus volume)
of the basins themselves. There are various uncertainties and
assumptions about the nature of these different contributions to
sea level which affect our confidence in projected sea-level
changes.
Projected changes in relative sea level from UKCP09 around
South-West England, under the UKCP09 medium emissions
scenario, are of the order of 0.45-0.5 m by 2090-2099, with
respect to a 1980-1999 baseline8. This range is that shown
around the coast of South-West England in figure 3.7 of Lowe
et al. (2009). The estimate quoted here combines the absolute
sea-level change estimates averaged around the UK (for the
central estimate, under the UKCP09 medium emissions scenario
only) and the vertical land movement resulting from glacial
isostatic adjustment since the last Ice Age. Regarding projected
future contributions to sea-level change from ice sheets, it is
assumed that recently-observed acceleration of glacial outflow
at ice-sheet edges will remain as recently observed, rather than
Climate change and geohazards in South-West England
ceasing or increasing. Changes in ice-sheet dynamics are
currently not well understood and there is considerable
uncertainty in this component of projected sea-level change. A
coastal flooding scenario for 2100, called ‘H++’, was also
included in UKCP09 (Lowe et al., 2009; Nicholls et al., 2011).
This scenario is based on faster rates of ice sheet melt. It was
designed to enable users to consider a more extreme outcome
than those in the main suite of projections, but one which is still
physically plausible (though with low probability). Only the
2100 time horizon was considered and so no time series is
given. This scenario will be discussed below together with the
other UKCP09 projections.
A plot of the projected change in relative sea level
throughout the 21st Century under all three UKCP09 emissions
scenarios is given for the example location of Penzance (Figure
2). There is relatively little variation in the projections for
locations around the coast of South-West England. Projected
changes for three other locations were assessed, but are not
shown for this reason. The 50th percentile (central estimate) of
the projected changes under the three emissions scenarios for
Penzance is around 0.4-0.6 m by the end of the century;
however, factoring in the uncertainties on these estimates
increases the projected range (5th-95th percentile range) to
around 0.2-0.9 m. Note that changes greater than this cannot
be ruled out as the H++ scenario range for time-mean sea-level
rise around the UK is 0.93-1.9 m. The top of this range is very
unlikely9 to occur in the 21st Century, but no further
quantification of the probability of occurrence of the H++
scenario is made (Lowe et al., 2009).
in the UK, the Netherlands and Belgium (Gerritsen, 2005), and
24,000 houses were damaged in the UK alone (Baxter, 2005).
In UKCP09, changes in storm surge have been considered in
terms of the ‘50-year return level’. This is the size of surge that
could be statistically expected to occur once per 50 years11. The
largest projected changes in storm surge around the UK are
found in the Bristol Channel and Severn Estuary, where the
trend is for an increase in the 50-year skew surge return level
of around 0.8 mm a–1 not including relative mean sea-level
change. However, overall it is suggested that the storm surge
component of extreme sea level will be much less than that
projected previously in UKCIP02. In terms of offshore wave
climate, seasonal mean and extreme waves are generally
projected to increase slightly to the south-west of the UK (Lowe
et al., 2009). There is large uncertainty in projected changes to
both storm surges and wave climate, however, and only a small
ensemble of climate model projections were used. The ranges
of the projected changes in these quantities should therefore be
considered as minimum ranges.
STRATEGIES FOR MANAGING PROJECTED CHANGES
TO THE CLIMATE
In view of the types of projected changes to the climate
which have been outlined above, what courses of action are
possible in order to manage the potential impacts of these
projected changes? There are two main approaches, namely
adaptation and mitigation.
Climate change adaptation is
defined as adjustment in natural and human systems in
response to actual or projected climate changes, which lessens
the effect of these changes or exploits any benefits (IPCC,
2007a). Examples of actions constituting adaptation include:
1)
building a new coastal power station on an elevated
platform in order to improve its resilience to sea-level rise;
2)
migration of a species within a particular ecosystem to a
new habitat, in response to the climate of the old habitat
becoming unsuitable for the species;
3)
choosing to grow crop varieties which are more
drought-tolerant in areas affected (or projected to be
affected) by water availability issues.
These examples show that adaptation can be ‘planned’ (i.e.
the result of a deliberate policy decision) or ‘autonomous’ (i.e.
not constituting a conscious response).
Climate change mitigation is defined as human intervention
which reduces the anthropogenic forcing of the climate system,
usually through reductions in GHG emissions and/or enhancing
GHG sinks (IPCC, 2007a). Examples of actions constituting
mitigation include:
Figure 2. Projected changes in relative sea level for Penzance,
under low, medium and high emissions scenarios. Solid lines
represent the 50th percentile values and the shaded areas around
each solid line represent the 5th–95th percentile range. Note that
the data are plotted w.r.t. 1990, as this is representative of the centre
of the 1980-1999 baseline period. Source: UKCP09.
Projected changes to other marine phenomena
around South-West England
As well as sea-level changes, projected changes to storm
surges and waves have also been assessed (Lowe et al., 2009).
Storm surges10 are short-lived, local increases in water level
above that of the tide, and are driven by weather (atmospheric
pressure gradients and winds). One of the most infamous surge
events in the UK was that which affected the coastline of the
southern North Sea in 1953, in which over 2,000 lives were lost
1)
improving domestic and industrial energy efficiency, so
that for a given amount of usable energy, a smaller
amount of fossil fuel is burned, and fewer GHGs are
emitted;
2)
making use of alternative energy sources where CO2
emissions are lower than those associated with burning
fossil fuels;
3)
making use of carbon capture and sequestration (CCS)
technologies at power generation facilities, so that
GHGs which would otherwise be emitted to the
atmosphere are captured in situ and subsequently
pumped into underground geological formations for
long-term storage;
4)
enhancing natural processes which remove GHGs from
the atmosphere, for example afforestation/reforestation
to enhance the CO2 sink associated with photosynthesis.
37
E.J. Palin
The following important points (Parry et al., 2007) about
adaptation and mitigation strategies should also be noted:
1)
Because a certain amount of warming of the climate is
already unavoidable, due to past GHG emissions,
adaptation is seen as a necessary action.
2)
Different temporal and spatial scales are relevant to
adaptation and mitigation respectively - mitigation has
global benefits but these may not be perceivable until
the middle of this century (as a result of the unavoidable
warming outlined in the preceding point). On the other
hand, adaptation is typically undertaken at local or
regional scales, but its benefits may be immediate.
3)
The best strategy for managing projected changes to the
climate is to undertake a mixture of both adaptation and
mitigation measures.
4)
There can be complex interactions between adaptation
and mitigation strategies. (For example, an adaptation
action to increased temperatures inside buildings is to
use air-conditioning. However, if the energy source for
this is derived from fossil-fuel sources, there is a
negative impact in terms of mitigation).
5)
The costs (both financial and otherwise) of some
adaptation and mitigation strategies mean that not all
actions are accessible to all societies.
CLIMATE CHANGE RISK AND
IN SOUTH-WEST ENGLAND
SELECTED GEOHAZARDS
The complex links between the various elements of the
Earth system include the potential for changes in the climate to
influence a selection of geological and geomorphological
hazards (Forster et al., 2009; Liggins et al., 2010). Many of the
links between climate and geohazards, such as those in
volcanic and mountainous regions, are not relevant for SouthWest England. However, others do have relevance, including
the links between:
1)
rainfall and flash flooding as evidenced by the Boscastle
and Lynton/Lynmouth flooding events;
2)
rainfall, local geology/geomorphology, and river
flooding in, for example, the risk of flooding in Exeter,
which is a key focus of the Environment Agency (EA)
(Environment Agency, 2012a);
3)
rainfall,
local
geology/geomorphology,
and
rockfalls/landslides in, for example, the rockfall and
landslide events at Burton Bradstock and Beaminster in
July 2012, which resulted in three fatalities (BBC, 2012;
British Geological Survey, 2012);
4)
sea-level change and impacts on the coastal zone.
The first and last of these examples are examined in the
following sections, through the use of case studies, which will
also be used to explain a selection of other challenges
associated with climate modelling and climate impacts research,
in particular the assessment of the climate change risk.
There are many definitions of the term ‘risk’; one common
definition is “risk = hazard × vulnerability” (e.g. Taubenböck
et al., 2008 and references therein). Here, ‘hazard’ describes the
projected change in a climatic parameter, typically defined as a
function of the probability of the change and the magnitude of
the change, and ‘vulnerability’ is a function of the exposure to
the change, resilience to the change and adaptive capacity
(ability to adapt to the change). In a risk assessment, metrics
of all the components of the hazard and vulnerability functions
38
may be assigned scores relative to each other, and a particular
risk may be quantified by multiplying the relevant hazard and
vulnerability scores together.
Often, the aim of undertaking a climate change risk
assessment is to inform future resilience planning in terms of
managing the risk. Risk may be addressed through adaptation
and/or mitigation; since the risk is location- and contextspecific, any strategy for reacting to a particular risk (or risks)
will be similarly specific.
However, using the above
framework, a climate change risk assessment can be used in
order to quantify and compare different risks in different
contexts and/or locations.
Describing the uncertainties
associated with making projections and assessing risk serves to
highlight the challenge of making decisions about the future,
such as those relevant to adaptation and/or mitigation, in the
face of uncertain information. Given the potential for nonlinear changes to the climate system and/or the possible future
occurrence of climatic variations outside the present realm of
human experience, it may become increasingly inappropriate to
assume that past conditions are fully representative of those
which might be seen in the future, and this is an additional
uncertainty to be considered.
Risk associated with sea-level rise
Projected rises in sea level pose a threat to the coastal
environment. An increase in average sea level can not only
result in direct inundation itself, but also increase the potential
impact of waves, storm surges, etc., since these would then be
superimposed on a higher background water level.
Projections for South-West England are for a relative sealevel rise of up to ~0.9 m (relative to 1980–1999) by the end of
this century, and a greater rise of up to 1.9 m under the H++
scenario (excluding storm surge and wave contributions in both
cases). What might this mean for the coast of the South-West?
An assessment of relative rock resistance across Great Britain
(Clayton and Shamoon, 1998; May and Hansom, 2003) found
that the coast of South-West England is generally less resistant
to erosion than that of either Scotland or Wales, but generally
more resistant than most of the rest of England. Considering
the South-West specifically, the assessment found that most of
Devon and Cornwall’s coast falls into the ‘low average’ or ‘high
average’ categories12. The main ‘resistant’ stretches are along
the North Devon coast and around the Penwith peninsula in
Cornwall. Most of the coast of Dorset, Somerset, Avon and
Gloucestershire is classed as ‘weak’ or ‘very weak’. A previous
study (Jones et al., 1988) identified five sections of the SouthWest England coast which were particularly prone to
landsliding (including parts of Weymouth Bay and Lyme Bay,
and the coast from Boscastle to the Taw Estuary). Four of these
five coastal sections are in areas identified as having ‘low
average’ or ‘weak’ resistance (the fifth is in an unclassified
area). However, it should be noted that the causes of these
coastal landslides are not necessarily related to erosion by the
sea. Furthermore, Jones et al. (1988) also stressed that
reporting of coastal landslides tended to be limited to those
which were conspicuous or which had implications for coastal
defence works.
Whatever an area’s degree of resistance to erosion, if it is
relatively low-lying and/or has critical infrastructure in the
coastal zone, it is potentially sensitive to sea-level rise. An
illustrative example is the coast near Dawlish in Devon. The
route of the main railway line linking London with Cornwall
and much of south Devon passes along this coast. The railway
is protected by sea defences, with the line frontage in this
vicinity being actively managed and maintained by Network
Rail and the EA (Dawson, 2012). However, severe weather can
still result in disruption to services, either directly or as a result
of remedial works following weather-related damage to the
defences. As well as this obvious logistical importance,
Dawlish Warren, a sandspit extending across the mouth of the
Exe estuary, is largely covered by a National Nature Reserve,
indicating a further ecological importance. The area is also a
Climate change and geohazards in South-West England
popular destination for tourism and recreation.
Figure 3 illustrates some key points of Dawlish Warren that
are relevant to the way in which the climate risk associated with
sea-level rise in this area might be assessed. Elevation data,
from the NEXTMap® Britain digital terrain model (DTM), are
shown on the figure13. The elevation intervals used have been
chosen to be an illustrative indicator of certain scenarios of
projected sea-level rise as follows: 0.5 m to represent a
moderate sea-level rise projection for the end of the century
(approximately the 50th percentile/central estimate value under
the UKCP09 medium emissions scenario, see Figure 2); 1 m to
represent a high sea-level rise projection for the end of the
century (greater than the 95th percentile value, ~0.9 m, under
the UKCP09 high emissions scenario, see Figure 2); and 1.9 m,
the upper end of the H++ scenario14. Hence, areas with
elevation below 0.5 m could be viewed as potentially sensitive
to sea-level rise of that amount. This is known as the ‘bathtub’
approximation, for obvious reasons. It is a rather crude
approach, in that it does not account for winds and tides,
erosional/accretional processes, or landform migration, and it
assumes no hydrological connectivity and no human
intervention (e.g. flood defences). However, in view of the
other uncertainties, including the fact that only the projected
mean sea-level rise component and not the projected changes
in storm surge is considered, it is an appropriate approximation
to give an indication of potentially sensitive areas.
Also included in Figure 3 are flood map data provided by the
EA (Environment Agency, 2006) showing present-day ‘extreme
flood’. This denotes a flood which has a 1-in-1000-year return
level (i.e. a 1-in-1000 probability of occurring during a
particular year) but does not differentiate between flooding
from rivers or the sea. Other relevant elements shown in Figure
3 are the railway line, areas of settlement, and the extent of the
National Nature Reserve.
In terms of a climate risk assessment, it would be useful to
use the types of information presented in Figure 3 to make
estimates of key elements of the risk posed to this area by
projected sea-level rise. For example, almost all of Dawlish
Warren, and areas adjacent to the railway line (though not
much of the line itself, as it generally has a higher elevation) are
in present-day ‘extreme flood’ zones, as are a few settlement
areas at the southern end of Dawlish Warren. Only one small
flood defence system, adjacent to the railway, is present in the
flood map database. However, the EA is still in the process of
adding existing defences to its database, so the sea defences
protecting the railway, between Dawlish and the southern end
of Dawlish Warren, are not yet included in its flood map, nor
are the defences along Dawlish Warren itself which include
gabion baskets, rock armour and timber groynes (Environment
Agency, 2012b).
A moderate sea-level rise, of the order of 0.5 m, has the
potential to threaten part of the Nature Reserve and the margins
of the sand spit itself. As the amount of sea-level rise increases,
so does the threatened area of the Nature Reserve. For the
most part, the elevation of the railway is greater than the upper
H++ value of 1.9 m, although there are areas further up the Exe
estuary (near Powderham Castle) where the railway has an
elevation of less than 1 m. A few settlement areas near the
Warren could be threatened by sea-level rise of the order of
1 m, and the number of potentially threatened settlement areas
rises if we consider sea-level rise of up to 1.9 m.
One of the challenges of understanding the risk posed to a
particular asset15 by climate change is to understand adequately
the hazard and vulnerability, even for the baseline period
against which future projected changes are measured.
Understanding these components of the baseline risk is an
essential step in being able to assess the projected future
changes in risk. For Dawlish Warren, the hazard is the
projected change in sea level. Although we have present-day
flood map information for this area, we do not know whether
the relevant driver is river or sea flooding, so we cannot assess
the contribution of sea flooding in the present day. In terms of
the future, a change in the hazard is a function of the
probability and the magnitude of the change. Hazard can be a
fairly complex function, with inherent uncertainty, and the
vulnerability can be even more difficult to define as it depends
on exposure, resilience and adaptive capacity. The exposure
could be considered in terms of a metric such as the percentage
of land in an area whose elevation is less than a certain value,
or the ‘value’ of infrastructure/assets - such as the settlement
areas, the railway, or the Nature Reserve - which lies inside the
‘extreme flood’ risk zone, or below a certain elevation.
(Consideration would be required as to how a ‘value’ could be
assigned to both the Nature Reserve and the railway so that
these ‘values’ are directly comparable. Typically the ‘value’ of
a natural system is quantified in terms of its ecological
importance, but the railway’s ‘value’ is of a more
financial/infrastructural nature).
The resilience might be
defined in terms of the quality and quantity of existing flood
defences - for example, the return level of flooding which the
flood defences in this area are built to resist. The adaptive
capacity could be the relative ease with which existing flood
defences can be modified to cope with the projected change in
hazard, or the extent to which a species living in the Nature
Reserve could tolerate (e.g. by migration) any loss of its habitat
as a result of the projected change in hazard.
All of these quantities require some effort in order to
understand and define them appropriately, which would be a
crucial early stage of any risk assessment. Nonetheless,
progress is being made in this type of research. Integrated
coastal models are being developed which incorporate many
processes in the ocean and atmosphere, and on land. Other
elements such as socio-economic factors may also be
incorporated. These models are used in various approaches to
coastal risk assessment, often using case studies in the UK (e.g.
Dawson et al., 2009; Robins et al., 2011). An example with
relevance to South-West England is the study by Purvis et al.
(2008) which proposed a framework for coastal flood risk
assessment that incorporates the uncertainty in future sea-level
rise projections, and applied it to a case study area
encompassing Weston-super-Mare, on the Somerset coast. The
study used an inundation model to examine the effects of
potential future extreme water levels and concluded that a risk
assessment based on a single (the ‘most plausible’) sea-level
rise scenario could underestimate financial losses, as this
neglected low-probability/high-impact scenarios. Further work
(Lewis et al., 2011) has built on this study, using the same
inundation model and incorporating UKCP09 data and thereby
allowing further quantification of uncertainties. Some attempt
was made to express the flood risk using an approach based on
financial losses, but the assessment was approximate. In
addition, although the H++ scenario was mentioned, it was not
used quantitatively in a risk assessment.
A further, more direct way in which this type of issue can be
addressed is to facilitate the sharing of information and
expertise between different organisations such as the Met
Office, EA, Ordnance Survey, British Geological Survey, etc., so
that an ‘all hazards’ approach (i.e. considering weather/climaterelated hazards along with other natural hazards) may be
followed. This is now beginning to be considered as part of the
‘Natural Hazards Partnership’ (NHP), a recently-formed
collaboration between the Met Office and twelve other partners
whose scope includes a variety of natural hazards. Although
the focus is on responding to hazards on the timescale of
weather rather than climate, the NHP will doubtless encourage
collaboration between experts from these organisations on
issues relevant to longer timescales. As such, there is hope that
the assessment of the hazard and vulnerability components of
climate risk will gradually become less complex. Additionally,
there are clear advantages of collaborative efforts between
natural/social scientists and stakeholders. For example, in a
recent initiative the EA worked with local residents to design
new flood mitigation measures in the Yorkshire town of
Pickering (Whatmore and Landström, 2011).
39
E.J. Palin
a
b
Figure 3. (a) The Exe estuary and (b) Dawlish Warren, showing a variety of local information relevant to a potential assessment of the
risk posed to the area by projected sea-level rise. NEXTMap® Britain elevation data (Intermap Technologies, 2012) are used for Dawlish
Warren to give a broad, illustrative indicator of areas which are potentially sensitive to projected sea-level rise of varying amounts. EA flood
map data - which are for present-day, not projected future, flood risk - are also included. The ‘extreme flood’ (green hatching) data are
based on a flood with a 1-in-1000-year return period.
40
Climate change and geohazards in South-West England
Risk associated with flash flooding
Examples of major flash flood events in South-West England,
which occur as a direct result of localised, very heavy rainfall,
include the Lynton and Lynmouth floods of 15th/16th August,
1952 (Delderfield, 1953) and the Boscastle floods of 16th
August, 2004 (HR Wallingford Ltd., 2005). Both of these events
were generated by periods of exceptionally heavy rainfall in a
relatively small area, with the Lynton and Lynmouth floods
claiming 34 lives (Delderfield, 1953). Rainfall totals for the 16th
August, 2004 across South-West England from the UK 5 km
gridded daily precipitation data set, which covers the period
1958-2009 (Perry et al., 2009), are shown in Figure 4a, while
the same data expressed as a percentage of the 1961-1990
climatological monthly (average total) rainfall for August are
presented in Figure 4b. The largest gridbox total in the
Boscastle area is 186.5 mm (cf. nearby observations of 200.4 mm
at Otterham and 184.9 mm at Trevalec), which is the secondhighest daily total in any gridbox in the region for the period
1958-2009. Around Boscastle, the daily rainfall totals on 16th
August, 2004 are of the order of 1-2 times that of the average
total August rainfall for 1961-1990.
Figure 5a shows observed and forecast rainfall rates for the
day of the Boscastle event. The event was associated with a
localised band of very heavy convective rainfall which persisted
for some hours over the same area. Figures 5b-d show the
effect of different resolutions of the forecast model on the
nature of the rainfall that was forecast. The 12 km model does
not identify any areas of heavy rainfall; the 4 km model predicts
some heavy rain, but does not correctly identify the shape of
the features. Only the 1 km model comes close to matching the
shape of the observed rainfall features (Roberts, 2006).
Unfortunately, at the time of the event, the operational model
was that with 12 km resolution, and the 4 km and 1 km cases
were run retrospectively to investigate their ability to reproduce
the meteorological features of this event.
The research challenge embodied by this case study is of a
more directly technical nature than that for Dawlish Warren
discussed above. Extreme events like the Boscastle and Lynton
and Lynmouth rainfall events are often the ones which are of
the greatest interest, because of their impacts. Yet climate
models, which typically have a far coarser resolution (25-300
km) than the forecast models just discussed, are unable to
represent convective rainfall extremes.
In consequence,
projection of future changes in flash flooding events of this
magnitude, using the present generation of climate models, is
currently not possible. Therefore, in order to draw any
conclusions about the potential effects of climate change on
flash flooding, it is necessary to find alternative approaches
which provide a useful, and scientifically robust, insight into the
subject. One such approach is to examine moderately high
percentiles of the rainfall distribution as represented by a
climate model. In so doing, it is assumed that changes in
moderately high percentiles, which climate models are better at
representing than more extreme events, give useful information
about changes in more extreme events. Figure 6 shows an
example using output from a perturbed physics ensemble of the
Met Office’s regional climate model, HadRM3.
This PPE contains 11 ensemble members, and is part of the
larger ensemble underlying UKCP09. Rainfall distribution for
the winters (December, January, February) and summers (June,
July, August) of the years 1961-1990 was assessed, and the
amount of rainfall corresponding to the 95th percentile at each
25 km gridbox was noted. The exceedance of this amount for
b
a
Figure 4. Rainfall amounts from the Met Office 5 km gridded observational data set on 16th August 2004 (Boscastle flooding event), expressed
(a) in mm and (b) as a percentage of the mean monthly rainfall (1961-1990) for August. Note non-linear scales in both figures.
a
b
c
d
Figure 5. Rainfall data from observations and forecasts for 16th August 2004, the day of the Boscastle floods. (a) Observed hourly rainfall
rates, from radar data, for 1500BST (BST = UTC+1). Forecast model output - forecast rainfall rates at 14 UTC from (b) 12 km forecast
starting at 00 UTC, (c) 4 km forecast starting at 01 UTC and (d) 1 km forecast starting at 01 UTC. Reproduced from Roberts (2006).
41
E.J. Palin
three future time horizons - the 2020s, 2050s and 2080s - was
then computed and is presented in Figure 6. The maps in the
rows of this figure represent the percentage change in
exceedance of the rainfall amount representing the wettest 5%
of days in the model’s baseline climate, at the three future time
horizons. The future time horizons have been defined in the
same way as those in UKCP09 (see above). Use of an ensemble
also allows the calculation of a minimum, mean and maximum
projected change across the ensemble; and these are shown in
the maps of the columns of Figure 6.
It is evident that the mean projected change is an increase in
exceedance for winter (Figure 6b), and a decrease in
exceedance for summer (Figure 6a). Thus, on average, the
occurrence of rainfall corresponding to the modelled 5% of
wettest winter days in 1961-1990 is projected to increase in
future, and rainfall corresponding to the modelled 5% of wettest
summer days in 1961-1990 is projected to decrease in future.
However, for both summer and winter rainfall, the modelled
range spans zero for all time periods (summer, Figure 6a) and
for the 2020s (winter, Figure 6b), so that there is uncertainty
even in the sign of the projected changes for these cases.
Conversely, an increase for winter rainfall in the 2050s and
2080s is robust across the ensemble.
However, the caveats and limitations of this analysis should
be noted. Firstly, regarding the extent to which various
uncertainties are sampled, an 11-member PPE run under the
medium emissions scenario does not sample structural
uncertainty (all 11 ensemble members are versions of the same
climate model), but it does sample parameter uncertainty to
some extent (different values of key parameters are used,
though this type of uncertainty is sampled more extensively in
UKCP09). Nor is emissions uncertainty sampled here, as only
one scenario has been used. Natural internal variability is
sampled to some extent. Changes have been quoted for three
different future periods, but with respect to only one baseline
period (and with no assessment of external variability).
Secondly, regarding the abilities of the model itself, projected
changes in 95th-percentile rainfall at the 25 km RCM grid scale
have been examined. This corresponds to heavy, but not
extreme, rainfall at a relatively large spatial scale.
Our
confidence in the ability of models to capture such events, and
hence projected changes in these, is relatively high. However,
it is less clear whether the projected changes to this type of
rainfall are indicative of what might be expected for more
extreme events. There are issues with the ability of current
climate models to capture the convective rainfall extremes
associated with summer storms, and research is ongoing (e.g.
Kendon et al., 2012) using higher-resolution models to look at
extreme rainfall processes and projected changes in more
detail. Thirdly, the rainfall amounts associated with the 95th
percentile have not been validated against observed rainfall
data. However, here we are using the top 5% of events in the
model as being representative of the top 5% of events in reality,
even if there is disagreement in their absolute amount. Finally,
the Boscastle event was a >99th percentile event, whereas the
95th percentile has been used here, because the statistics of
exceedance of the 99th percentile are less robust, since it is
only exceeded 1% of the time. The 95th percentile is an
appropriate compromise involving better statistics of
exceedance, but still a relatively heavy event.
UKCP09 includes a “change in precipitation on the wettest
day of the season” variable, which (given the length of a season
is ~100 days) corresponds approximately to the 99th percentile
of the season. We can compare the PPE results with projections
for this variable, although it must be borne in mind that (a) the
results are for different percentiles of the rainfall distribution
and (b) the range in the PPE projections does not directly
translate into a range across UKCP09 probability levels. The
best approach is to compare the mean PPE response for the
95th percentile with the 50% probability level (central estimate)
of the UKCP09 wettest day change, under the medium scenario
for consistency. This shows that the sign of the change
(positive) is consistent between the two for winter, for all three
42
future time periods. The situation is less clear for summer. The
UKCP09 wettest day change at the 50% level across South-West
England can be positive or negative in different 25 km grid
squares, but the mean PPE response is negative for all three
time periods. The differences between UKCP09 and the 11member PPE for summer rainfall can be explained by the fact
that the full UKCP09 ensemble samples more uncertainty.
As is the case for the sea-level rise example, the projected
future risk posed by flash flooding events is a complex function
of both the hazard of, and vulnerability to, these events. For
Boscastle, and Lynton and Lynmouth, geological and
geomorphological aspects are evident in the vulnerability.
Boscastle and the River Valency catchment within which it lies
are underlain by sedimentary rocks of late Devonian to early
Carboniferous age (British Geological Survey, 2011a, b).
Lynton and Lynmouth are underlain by early-to-mid Devonian
mudstone, siltstone and sandstones of the Lynton Formation
while the wider River Lyn catchment is mostly underlain by
mid-Devonian interbedded sandstone and conglomerates of the
Hangman Sandstone Formation (British Geological Survey,
2011a, b). The geomorphology of both areas is similar, with
deep, steep-sided river valleys. For both events, the antecedent
weather conditions had been relatively wet (Delderfield, 1953;
Doe, 2004), leading to increased saturation of the land and thus
an increased likelihood that further rainfall would lead to runoff
rather than drainage through the soil, especially since the soils
in these two areas are comparatively shallow.
Other
contributing factors included the damming of the rivers by
debris being carried downstream, and the subsequent failure of
these temporary dams resulting in sudden surges of water along
the path of the rivers (particularly in the case of the Lynton and
Lynmouth event).
As well as emphasising again that assessing climate risk can
be a complex process, this case study has highlighted the
further challenge that there are important aspects of the climate
which are difficult or impossible to assess using the current
generation of climate models. This can be addressed by
continuing to push forward the development of models, but the
timescales involved mean that alternative ways to assess some
of these issues using current models must be found. As a result,
there will always be caveats and limitations associated with a
given approach, which must be clearly stated along with the
results of any analysis.
SUMMARY
AND CONCLUSIONS
This paper has provided a brief overview of the main causes,
both natural and anthropogenic, of (geologically) recent global
climate change. Climate change projections at the global, UK
and regional scale have been discussed, and UKCP09, the UK’s
most recent government-funded suite of climate change
projections, has been used to give a summary of the projected
changes in temperature, precipitation and sea level rise for
South-West England in the coming decades, together with
commentary on the differing levels of confidence in the
projected changes of different variables.
The types of uncertainty in climate projections have been
considered, together with ways in which these can be
managed, and in some cases quantified and/or reduced. A brief
overview of the types of strategies for managing projected
changes to the climate, namely, climate change adaptation and
mitigation, has been presented. However, for a given asset
(whether it be a location, organisation, ecosystem, etc.), climate
change projections are only part of the story in assessing the
potential risk posed by climate change to that asset.
Information is required not only about the projected evolution
of a climate-related quantity (and thus the changing ‘hazard’),
but also about what impacts the change in hazard might have
(and thus the changing ‘vulnerability’). It is also clear from this
that the hazard and vulnerability for the baseline period must
also be well understood, in order to make a meaningful
assessment of the potential changes in these. Finally, in order
Climate change and geohazards in South-West England
a
b
Figure 6. Projected changes in exceedance of 95th percentile (a) summer and (b) winter rainfall amounts in South-West England,
calculated from an 11-member PPE of HadRM3, relative to a baseline of 1961-1990.
43
E.J. Palin
to bring together some of the concepts covered in this paper,
some of the challenges associated with assessing climate risk (in
terms of both the hazard and vulnerability components) have
been explored, using two examples with both geological and
local relevance.
In summary, it remains a crucially important task for those
working in the fields of climate science and climate impacts
research to ensure that the science of climate change, the
potential impacts of climate change, and the consequences of
these impacts for society are assessed as effectively as possible,
and equally that the findings of these research areas are
communicated appropriately to stakeholders and the general
public.
NOTES
1
Here, ‘very likely’ and ‘likely’ are IPCC terminology, with
particular definitions in terms of probability (given in the
brackets).
Lorenz’s analogy in this reference is to a seagull’s wings - it
was not until the early 1970s that the butterfly metaphor arose
(Hilborn, 2004).
2
3
The link between orbital cycles and climate was proposed
previously by James Croll (Croll, 1864), a largely self-taught
scientist whose work as a museum caretaker allowed him
access to the books which helped him to develop his ideas - so
we might more properly refer to the cycles as ‘CrollMilankovitch cycles’.
The various known aerosol effects are summarised in figure
2.10 of Forster et al. (2007).
4
A further complication is that humans have an influence on
the extent of deserts, because of anthropogenic land-use
change, so not all desert-derived dust aerosol can be
considered as being natural in origin.
5
6
The IPCC defines ‘radiative forcing’ as “the change in the net
(downward minus upward) irradiance (expressed in W.m–2) at
the tropopause due to a change in an external driver of climate
change, such as, for example, a change in the concentration of
carbon dioxide or the output of the Sun. […]”. The full
definition may be found in IPCC (2007b), Annex I.
7
The ‘baseline’ climate is that against which projected future
changes are measured. It is often, though not always, taken to
be a period broadly representative of ‘present’ conditions.
When a climate parameter is arithmetically averaged over a 30year period, the resulting value is sometimes termed a ‘climate
normal’ (WMO, 2012).
Definitions of the baseline and future periods differ for sea
level rise projections from those for the other variables. This is
because the UKCP09 sea level rise projections are based on
IPCC projections and these time periods were used in the IPCC
assessment (Meehl et al., 2007).
assessment, so some stretches of coast are not represented in
the discussion presented here.
13
Since these DTM data are being discussed here in conjunction
with UKCP09 sea level rise projections, it should be noted that
the vertical datum for the DTM is ODN (Ordnance Datum
Newlyn), the Ordnance Survey’s national coordinate system for
heights above mean sea level, for which the time period is May
1915–April 1921. Some sea level rise will therefore have
occurred between this time and the baseline period for UKCP09
sea level rise projections, which is 1980–1999. A value of 13 cm
has been estimated for this discrepancy, using Newlyn tidal
gauge data (Permanent Service for Mean Sea Level (PSMSL),
2012; Woodworth and Player, 2003). It has been assumed that
this correction, valid for Newlyn, also applies to Dawlish (P.
Woodworth pers. comm.). For simplicity, the correction has not
been applied to the data shown on Figure 3, as its magnitude
is likely smaller than other uncertainties on the various data
shown in this figure.
14
It has been assumed that the vertical error in DTM elevations
allows one to distinguish adequately between these categories.
15
The word ‘asset’ is used in a very general sense to represent
any phenomenon or entity which may be affected by climate
change. Examples of ‘assets’ could include a business, an item
of infrastructure, a biological species, etc.
ACKNOWLEDGEMENTS
The author is grateful to Richard Betts, Jason Lowe and
Karina Williams for commenting on drafts of the manuscript,
Elizabeth Kendon for useful discussions about extreme rainfall
events and for commenting on drafts of the manuscript, John
Caesar for assistance with use of the gridded observational data
shown in Figure 4, Nigel Roberts for providing Figure 5, Philip
Woodworth (Proudman Oceanographic Laboratory) for advice
on interpreting tidal gauge data, and Jason Jones (Intermap) for
assistance with DTM data.
UKCP09 content in this paper is © Crown Copyright 2009.
The UK Climate Projections (UKCP09) have been made
available by the Department for Environment, Food and Rural
Affairs (Defra) and the Department of Energy and Climate
Change (DECC) under licence from the Met Office, UK Climate
Impacts Programme, British Atmospheric Data Centre,
Newcastle University, University of East Anglia, Environment
Agency, Tyndall Centre and Proudman Oceanographic
Laboratory. These organisations give no warranties, express or
implied, as to the accuracy of the UKCP09 and do not accept
any liability for loss or damage, which may arise from reliance
upon the UKCP09 and any use of the UKCP09 is undertaken
entirely at the user’s risk.
8
The use of ‘very unlikely’ in connection with H++ is not in line
with the IPCC definitions of similar phrases discussed elsewhere
in this paper.
9
10
More formally we should use the term ‘positive storm surges’,
as negative storm surges do occur under some conditions.
11
The return level is a purely statistical quantity. If a surge
greater than the 50-year return level occurs once, that does not
mean it will not then occur for another 50 years, nor must a
surge with 50-year return level occur exactly every 50 years.
12
There were some areas which are unclassified in the original
44
REFERENCES
BAXTER, P. 2005. The east coast Big Flood, 31 January-1 February 1953:
a summary of the human disaster. Philosophical Transactions of the Royal
Society A, 363, 1293-1312.
BRITISH BROADCASTING CORPORATION (BBC) 2012. Two people dead in car
buried by tunnel landslide. http://www.bbc.co.uk/news/uk-18876806.
BRITISH GEOLOGICAL SURVEY
2011a.
Geology of Britain viewer.
http://maps.bgs.ac.uk/geologyviewer_google/googleviewer.html.
BRITISH GEOLOGICAL SURVEY
http://www.bgs.ac.uk/lexicon/.
2011b.
Lexicon of Named Rock Units.
BRITISH GEOLOGICAL SURVEY 2012. Burton Bradstock rock fall, Dorset.
http://www.bgs.ac.uk/landslides/burtonBradstock.html.
Climate change and geohazards in South-West England
BINDOFF, N., WILLEBRAND, K., ARTALE, V., CAZENAVE, A., GREGORY, J.,
GULEV, S., HANAWA, K., LEQUÉRÉ, C., LEVITUS, S., NOJIRI, Y., SHUM, C.,
TALLEY, L. and UNNIKRISHNAN, A. 2007. Observations: Oceanic climate
change and sea level. In: SOLOMON, S., QIN, D., MANNING, M., CHEN, Z.,
MARQUIS, M., AVERYT, K., TIGNOR, M. and MILLER, H. (Eds), Climate
Change 2007: The physical science basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge University Press, Cambridge.
HART, M. and HART, A. 2000. Global climate change: a geological perspective.
Geoscience in South-West England, 10, 14-17.
HAWKINS, E. and SUTTON, R. 2009. The potential to narrow uncertainty in
regional climate predictions. Bulletin of the American Meteorological Society,
90, 1095-1107.
HAYS, J., IMBRIE, J. and SHACKLETON, N. 1976. Variations in the Earth’s orbit:
pacemaker of the Ice Ages. Science, 194, 1121-1132.
CLAYTON, K. and SHAMOON, N. 1998. New approach to the relief of Great
Britain II. A classification of rocks based on relative resistance to denudation.
Geomorphology, 25, 155–171.
HELD, I. 2005. The gap between simulation and understanding in climate
modeling. Bulletin of the American Meteorological Society, 86, 1609-1614.
CROLL, J. 1864. On the physical cause of the change of climate during geological
epochs. Philosophical Magazine, 28, 121-137.
HILBORN, R. 2004. Sea gulls, butterflies and grasshoppers: a brief history of the
butterfly effect in nonlinear dynamics. American Journal of Physics, 72, 425-426.
DAWSON, D. 2012. The impact of sea-level rise on the London-Penzance railway
line. Unpublished PhD thesis, University of Plymouth.
HR WALLINGFORD LTD. 2005. Flooding in Boscastle and North Cornwall,
August 2004. Phase 2 Studies Report. Technical Report EX 5160, HR
Wallingford Ltd.
DAWSON, R., DICKSON, M., NICHOLLS, R., HALL, J., WALKDEN, M., STANSBY, P.,
MOKRECH, M., RICHARDS, J., ZHOU, J., MILLIGAN, J., JORDAN, A.,
PEARSON, S., REES, J., BATES, P., KOUKOULAS, S. and WATKINSON, A.
2009. Integrated analysis of risks of coastal flooding and cliff erosion under
scenarios of long term change. Climatic Change, 95, 249-288.
DEPARTMENT FOR ENVIRONMENT, FOOD and RURAL AFFAIRS (DEFRA) 2011.
UK Climate Projections. http://ukclimateprojections.defra.gov.uk/.
DELDERFIELD, E. 1953. The Lynmouth flood disaster. E.R.D. Publications Ltd,
Exmouth.
DOE, R. 2004. Extreme precipitation and run-off induced flash flooding at
Boscastle, Cornwall, UK - 16 August 2004. Journal of Meteorology, 29, 319-333.
ENVIRONMENT AGENCY 2006. Understanding flood risk. Using our flood map:
identifying and understanding flood risk in England and Wales.
http://publications.environment-agency.gov.uk/pdf/GEH00306BKIY-e-e.
pdf?lang=_e.
ENVIRONMENT AGENCY 2012a. Exe Catchment Flood Management Plan:
Summary report June 2012. http://www.environmentagency.gov.uk/static/
documents/Leisure/_CFMP_Exe_2012.pdf.
ENVIRONMENT AGENCY 2012b Repairing the ‘backbone’ of Dawlish Warren.
http://www.environment-agency.gov.uk/news/135992.aspx.
EPICA COMMUNITY MEMBERS 2004. Eight glacial cycles from an Antarctic ice
core. Nature, 429, 623-628.
FORSTER, A., CULSHAW, M., WILDMAN, G. and HARRISON, M.
2009.
Implications of climate change for urban areas in the UK from an engineering
geological perspective. In: CULSHAW, M., REEVES, H., JEFFERSON, I. and
SPINK, T. (Eds), Engineering geology for tomorrow’s cities. Engineering
Geology Special Publications, Geological Society, London 22, Paper 246.
FORSTER, P., RAMASWAMY, V., ARTAXO, P., BERNTSEN, T., BETTS, R., FAHEY, D.,
HAYWOOD, J., LEAN, J., LOWE, D., MYRHE, G., NGANGA, J., PRINN, R.,
RAGA, G., SCHULZ, M. and VAN DORLAND, R. 2007. Changes in
atmospheric constituents and in radiative forcing. In: SOLOMON, S., QIN, D.,
MANNING, M., CHEN, Z., MARQUIS, M., AVERYT, K., TIGNOR, M. and
MILLER, H. (Eds), Climate Change 2007: The physical science basis.
Contribution of Working Group I to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge.
FOUKAL, P., FRÖHLICH, C., SPRUIT, H. and WIGLEY, T. 2006. Variations in solar
luminosity and their effect on the Earth’s climate. Nature, 443, 161-166.
FRIEDLINGSTEIN, P., COX, P.M., BETTS, R.A., BOPP, L., VON BLOH, W.,
BROVKIN, V., CADULE, P., DONEY, S., EBY, M., FUNG, I., BALA, G., JOHN, J.,
JONES, C.D., JOOS, F., KATO, T., KAWAMIYA, M., KNORR, W., LINDSAY, K.,
MATTHEWS, H.D., RADDATZ, T., RAYNER, P., REICK, C., ROECKNER, E.,
SCHNITZLER, K.G., SCHNUR, R., STRASSMAN, K., WEAVER, A.J.,
YOSHIKAWA, C. and ZENG, N. 2006. Climate-carbon cycle feedback analysis:
Results from the C4MIP model intercomparison. Journal of Climate, 19,
3337-3353.
GERRITSEN, H. 2005. What happened in 1953? The Big Flood in the Netherlands
in retrospect.
Philosophical Transactions of the Royal Society A, 363,
1271-1291.
GORDON, C., COOPER, C., SENIOR, C., BANKS, H., GREGORY, J., JOHNS, T.,
MITCHELL, J. and WOOD, R. 2000. The simulation of SST, sea ice extents
and ocean heat transports in a version of the Hadley Centre coupled model
without flux adjustments. Climate Dynamics, 16, 147-168.
INESON, S., SCAIFE, A., KNIGHT, J., MANNERS, J., DUNSTONE, N., GRAY, L. and
HAIGH, J. 2011. Solar forcing of winter climate variability in the northern
hemisphere. Nature Geoscience, 4, 753-757.
INTERMAP TECHNOLOGIES 2012. NEXTMap products: digital terrain models,
digital surface models, orthorectified radar images, GIS maps, 3D terrain.
http://www.intermap.com/en-us/solutions/nextmap.aspx.
IPCC 2007a. Climate Change 2007: Impacts, adaptation, and vulnerability.
Contribution of Working Group II to the Fourth Assessment Report of the
Intergovernmental Panel on Climate Change. Cambridge University Press,
Cambridge.
IPCC 2007b. Climate Change 2007: The physical science basis. Contribution of
Working Group I to the Fourth Assessment Report of the Intergovernmental
Panel on Climate Change. Cambridge University Press, Cambridge.
IPCC 2007c. Summary for policymakers. In: SOLOMON, S., QIN, D., MANNING, M.,
CHEN, Z., MARQUIS, M., AVERYT, K., TIGNOR, M. and MILLER, H. (Eds),
Climate Change 2007: The physical science basis. Contribution of Working
Group I to the Fourth Assessment Report of the Intergovernmental Panel on
Climate Change. Cambridge University Press, Cambridge.
JANSEN, E., OVERPECK, J., BRIFFA, K., DUPLESSY, J.-C., JOOS, F., MASSON
DELMOTTE, V., OLAGO, D., OTTO-BLIESNER, B., PELTIER, W., RAHMSTORF, S.,
RAMESH, R., RAYNAUD, D., RIND, D., SOLOMINA, O., VILLALBA, R. and
ZHANG, D. 2007. Palaeoclimate. In: SOLOMON, S., QIN, D., MANNING, M.,
CHEN, Z., MARQUIS, M., AVERYT, K., TIGNOR, M. and MILLER, H. (Eds),
Climate Change 2007: The physical science basis. Contribution of Working
Group I to the Fourth Assessment Report of the Intergovernmental Panel on
Climate Change. Cambridge University Press, Cambridge.
JENKINS, G., MURPHY, J., SEXTON, D., LOWE, J., JONES, P. and KILSBY, C. 2009.
UK Climate Projections: briefing report. Met Office Hadley Centre, Exeter.
JONES, C.D., HUGHES, J.K., BELLOUIN, N., HARDIMAN, S.C., JONES, G.S.,
KNIGHT, J., LIDDICOAT, S., O’CONNOR, F.M., ANDRES, R.J., BELL, C., BOO,
K.-O., BOZZO, A., BUTCHART, N., CADULE, P., CORBIN, K.D., DOUTRIAUX
BOUCHER, M., FRIEDLINGSTEIN, P., GORNALL, J., GRAY, L., HALLORAN,
P.R., HURTT, G., INGRAM, W.J., LAMARQUE, J.-F., LAW, R.M., MEINSHAUSEN,
M., OSPREY, S., PALIN, E.J., CHINI, L.P., RADDATZ, T., SANDERSON, M.G.,
SELLAR, A.A., SCHURER, A., VALDES, P., WOOD, N., WOODWARD, S.,
YOSHIOKA, M. and ZERROUKAT, M.
2011.
The HadGEM2-ES
implementation of CMIP5 centennial simulations.
Geoscientific Model
Development, 4, 543-570.
JONES, D., GRIFFITHS, J. and LEE, E. 1988. The distribution of recorded landslides
in south-west England. Proceedings of the Ussher Society, 7, 91-92.
JONES, R., MURPHY, J., NOGUER, M. and KEEN, A. 1997. Simulation of climate
change over Europe using a nested regional-climate model. II: Comparison of
driving and regional model responses to a doubling of carbon dioxide
concentration.
Quarterly Journal of the Royal Meteorological Society,
123, 265-292.
KENDON, E., ROBERTS, N., SENIOR, C. and ROBERTS, M. 2012. Realism of
rainfall in a very high resolution climate model. Journal of Climate, 25,
5791-5806.
LE TREUT, H., SOMERVILLE, R., CUBASCH, U., DING, Y., MAURITZEN, C.,
MOKSSIT, A., PETERSON, T. and PRATHER, M. 2007. Historical overview of
climate change science. In: SOLOMON, S., QIN, D., MANNING, M., CHEN, Z.,
MARQUIS, M., AVERYT, K., TIGNOR, M. and MILLER, H. (Eds), Climate
Change 2007: The physical science basis. Contribution of Working Group I to
the Fourth Assessment Report of the Intergovernmental Panel on Climate
Change. Cambridge University Press, Cambridge.
45
E.J. Palin
LEWIS, M., HORSBURGH, K., BATES, P. and SMITH, R. 2011. Quantifying the
uncertainty in future coastal flood risk estimates for the UK. Journal of
Coastal Research, 27, 870-881.
PURVIS, M., BATES, P. and HAYES, C. 2008. A probabilistic methodology to
estimate future coastal flood risk due to sea level rise. Coastal Engineering,
55, 1062-1073.
LIGGINS, F., BETTS, R. and MCGUIRE, B. 2010. Projected future climate changes
in the context of geological and geomorphological hazards. Philosophical
Transactions of the Royal Society A, 368, 2347-2367.
ROBERTS, N. 2006. Simulations of extreme rainfall events using the Unified
Model with a grid spacing of 12, 4 and 1km. Forecasting technical report 486,
Met Office, Exeter.
LOCKWOOD, M. and FRÖHLICH, C. 2007. Recent oppositely directed trends in
solar climate forcings and the global mean surface air temperature.
Proceedings of the Royal Society A, 463, 2447-2460.
ROBINS, P., DAVIES, A. and JONES, R. 2011. Application of a coastal model to
simulate present and future inundation and aid coastal management. Journal
of Coastal Conservation, 15, 1-14.
LORENZ, E. 1963. The predictability of hydrodynamic flow. Transactions of the
New York Academy of Sciences, 25, 409-432.
ROYAL SOCIETY 2009. Geoengineering the climate.
Geoengineering-the-climate/.
LOWE, J., HOWARD, T., PARDAENS, A., TINKER, J., HOLT, J., WAKELIN, S.,
MILNE, G., LEAKE, J., WOLF, J., HORSBURGH, K., REEDER, T., JENKINS, G.,
RIDLEY, J., DYE, S. and BRADLEY, S. 2009. UK Climate Projections science
report: marine and coastal projections. Met Office Hadley Centre, Exeter.
SCAFETTA, N. and WEST, B. 2006. Phenomenological solar contribution to the
1900–2000 global surface warming. Geophysical Research Letters, 33, L05708.
MAY, V. and HANSOM, J. 2003. Coastal Geomorphology of Great Britain.
Geological Conservation Review Series 28. Joint Nature Conservation
Committee, Peterborough.
MEEHL, G., STOCKER, T., COLLINS, W., FRIEDLINGSTEIN, P., GAYE, A.,
GREGORY, J., KITOH, A., KNUTTI, R., MURPHY, J., NODA, A., RAPER, S.,
WATTERSON, I., WEAVER, A. and ZHAO, Z.-C. 2007. Global climate
projections. In: SOLOMON, S., QIN, D., MANNING, M., CHEN, Z., MARQUIS,
M., AVERYT, K., TIGNOR, M. and MILLER, H. (Eds), Climate Change 2007:
The physical science basis.
Contribution of Working Group I to the
Fourth Assessment Report of the Intergovernmental Panel on Climate Change.
Cambridge University Press, Cambridge.
MILANKOVIĆ, M. 1941. Kanon der Erdbestrahlung und seine Anwendung auf
das Eiszeitenproblem. Königlich Serbische Akademie, Special Publication
133, Belgrade. Translated as “Canon of Insolation of the Ice-Age Problem” by
the Israel Program for Scientific Translations, Jerusalem, 1969.
MURPHY, J., SEXTON, D., JENKINS, G., BOORMAN, P., BOOTH, B., BROWN, C.,
CLARK, R., COLLINS, M., HARRIS, G., KENDON, E., BETTS, R., BROWN, S.,
HOWARD, T., HUMPHREY, K., MCCARTHY, M., MCDONALD, R., STEPHENS,
A., WALLACE, C., WARREN, R., WILBY, R. and WOOD, R. 2009. UK Climate
Projections Science Report: Climate change projections. Met Office Hadley
Centre, Exeter.
NAKIĆENOVIĆ, N., ALCAMO, J., DAVID, G., DE VRIES, B., FENHANN, J.,
GAFFIN, S., GREGORY, K., GRÜBLER, A., JUNG, T.Y., KRAM, T., ROVERE,
E.L.L., MICHAELIS, L., MORI, S., MORITA, T., PEPPER, W., PITCHER, H.,
PRICE, L., RIAHI, K., ROEHRL, A., ROGNER, H., SANKOVSKI, A.,
SCHLESINGER, M., SHUKLA, P., SMITH, S., SWART, R., VAN ROOIJEN, S.,
VICTOR, N. and DADI, Z. 2000. IPCC Special Report on Emissions Scenarios.
Cambridge University Press, Cambridge.
NICHOLLS, R., HANSON, S., LOWE, J., WARRICK, R., LU, X., LONG, A. and
CARTER, T. 2011. Constructing sea-level scenarios for impact and adaptation
assessment of coastal areas: A guidance document. Supporting Material,
Intergovernmental Panel on Climate Change Task Group on Data and
Scenario Support for Impact and Climate Analysis (TGICA).
PARKER, D., WILSON, H., JONES, P., CHRISTY, J. and FOLLAND, C. 1996. The
impact of Mount Pinatubo on world-wide temperatures. International
Journal of Climatology, 16, 487-497.
PARRY, M., CANZIANI, O., PALUTIKOF, J. and CO-AUTHORS. 2007. Technical
summary. In: PARRY, M., CANZIANI, O., PALUTIKOF, J., VAN DER LINDEN,
P. and HANSON, C. (Eds), Climate Change 2007: Impacts, adaptation, and
vulnerability. Contribution of Working Group II to the Fourth Assessment
Report of the Intergovernmental Panel on Climate Change. Cambridge
University Press, Cambridge.
PENNELL, C. and REICHLER, T. 2011. On the effective number of climate
models. Journal of Climate, 24, 2358-2367.
PERMANENT SERVICE FOR MEAN SEA LEVEL (PSMSL). 2012. Tide gauge data
(Newlyn).
Retrieved 9th May 2012 from http://www.psmsl.org/data/
obtaining/stations/202.php.
PERRY, M., HOLLIS, D. and ELMS, M. 2009. The generation of daily gridded
datasets for temperature and rainfall for the UK. Climate Memorandum No.
24, National Climate Information Centre.
http://www.metoffice.gov.uk/
climatechange/science/downloads/generation_of_daily_gridded_datasets.pdf.
PETIT, J., JOUZEL, J.R.D., BARKOV, N., BARNOLA, J.-M., BASILE, I., BENDER, M.,
CHAPPELLAZ, J., DAVIS, M., DELAYGUE, G., DELMOTTE, M., KOTLYAKOV, V.,
LEGRAND, M., LIPENKOV, V., LORIUS, C., PEPIN, L., RITZ, C., SALTZMAN, E.
and STIEVENARD, M. 1999. Climate and atmospheric history of the past
420,000 years from the Vostok ice core, Antarctica. Nature, 399, 429-436.
46
http://royalsociety.org/
SCAIFE, A.A., COPSEY, D., GORDON, C., HARRIS, C., HINTON, T., KEELEY, S.,
O'NEILL, A., ROBERTS, M. and WILLIAMS, K. 2011. Improved Atlantic winter
blocking in a climate model. Geophysical Review Letters, 38, L23703.
SMITH, D., CUSACK, S., COLMAN, A., FOLLAND, C., HARRIS, G. and MURPHY, J.
2007. Improved surface temperature prediction for the coming decade from
a global climate model. Science, 317, 796-799.
SOLOMON, S., QIN, D., MANNING, M., ALLEY, R., BERNTSEN, T., BINDOFF, N.,
CHEN, Z., CHIDTHAISONG, A., GREGORY, J., HEGERL, G., HEIMANN, M.,
HEWITSON, B., HOSKINS, B., JOOS, F., JOUZEL, J., KATTSOV, V., LOHMANN,
U., MATSUNO, T., MOLINA, M., NICHOLLS, N., OVERPECK, J., RAGA, G.,
RAMASWAMY, V., REN, J., RUSTICUCCI, M., SOMERVILLE, R., STOCKER, T.,
WHETTON, P., WOOD, R. and WRATT, D. 2007. Technical summary. In:
SOLOMON, S., QIN, D., MANNING, M., CHEN, Z., MARQUIS, M., AVERYT, K.,
TIGNOR, M. and MILLER, H. (Eds), Climate Change 2007: The physical
science basis. Contribution of Working Group I to the Fourth Assessment
Report of the Intergovernmental Panel on Climate Change. Cambridge
University Press, Cambridge.
TAUBENBÖCK, H., POST, J., ROTH, A., ZOSSEDER, K., STRUNZ, G. and DECH, S.
2008. A conceptual vulnerability and risk framework as outline to identify
capabilities of remote sensing. Natural Hazards and Earth System Sciences,
8, 409-420.
TEBALDI, C. and KNUTTI, R. 2007. The use of the multi-model ensemble in
probabilistic climate projections. Philosophical Transactions of the Royal
Society A, 365, 2053–2075.
UKCP09 2012. Quick downloads. Technical notes can be accessed via
http://ukclimateprojections.defra.gov.uk/22544/.
VAQUERO, J. 2007. Historical sunspot observations: A review. Advances in
Space Research, 40, 929-941.
WHATMORE, S. and LANDSTRÖM, C. 2011. Flood apprentices: an exercise in
making things public. Economy and Society, 40, 582-610.
WITTMANN, A. and XU, Z. 1987. A catalogue of sunspot observations from 165
BC to AD 1684. Astronomy & Astrophysics Supplement Series, 70, 83-94.
WMO
2012.
Climate
data
and
data
related
products.
http://www.wmo.int/pages/themes/climate/climate_data_ and_products.php.
WOODWORTH, P. and PLAYER, R. 2003. The Permanent Service for Mean Sea
Level: an update to the 21st century. Journal of Coastal Research, 19, 287-295.